Feedback less PushPull Converter

[adnan012]

hi,

Is it possible to run a push-pull converter without any voltage feedback? With only current sense resistor at the primary side, can we get voltage regulation?

Regards

 

[codemaster11]

How it is possible to get voltage regulation without voltage feedback to the mosfet driver section.

the output will certainly drop by putting load on it.

 

[adnan012]

Thanks for reply.

Can i say that at zero load the Pulse width will be at its minimum value and at different loads PWM will not remain same and it will change. Or PWM will remain same in any case?

Regards

 

[codemaster11]

the duty cycle of the pulse will certainly change by putting load if the circuit has feedback.

in case if there is no feedback to the mosfet driver section the PWM duty cycle will not increase

cause the output voltage drop. PWM duty cycle goes to maximum value when there is feedback from

output to primary side driver section. use tl494 or sg3525 mosfet driver along with tl431 feedback block.

 

[schmitt trigger]

You need a bare-minimum type of feedback.

Something that I have seen, for very low cost applications, is an on-off feedback.

Essentially, you run the push pull at maximum (50-50) duty cycle. On the secondary, you only have a comparator measuring the voltage, and when it is exceeded, it sends a signal back to fully turn off the PWM.

The results in a larger ripple voltage, but if the system tolerates this, or you are doing post-regulation, it is a simple control scheme.

The only caveat is that you must ensure that when you turn on again the PWM, the driver must be the opposite of the driver that was energized when the PWM stopped. Otherwise you risk saturating the transformer and damaging the transistors.

 

[asdf44]

You should be able to run the forward converter 'open loop' with no feedback. Many switching supplies operate this way. But you'll get 0 line regulation and poor load regulation. Maybe this is ok especially if a following stage provides regulation.


If you're looking for cost cutting tricks for eliminating the opto-coupler look for examples like this flyback controller:
https://www.maximintegrated.com/en/products/power/isolated-power/MAX17690.html

 

[Easy peasy]

the only way to operate under the inference that Vo = D. Vin ( for 1:1:1 turns ratio ) is to use heavy subbers on the mosfets and o/p diodes - this will give you ideal text book waveforms and will not allow the volts to lift at no load - however the efficiency will suffer ... the load reg will then be a function of the wiring resistances, mosfet R, diode drop and of course the (approx) fixed snubber losses